Review

Spray and combustion characterization for internal combustion engines using

optical measuring techniques e A review
S.N. Soid
a, b
, Z.A. Zainal
a,
*
a
School of Mechanical Engineering, Universiti Sains Malaysia, Engineering Campus, 14300 Nibong Tebal, Penang, Malaysia
b
Mechanical Section, Universiti Kuala Lumpur Malaysian Spanish Institute, 09000 Kulim Hi-Tech Park, Kulim, Kedah, Malaysia
a r t i c l e i n f o
Article history:
Received 7 June 2010
Received in revised form
14 November 2010
Accepted 16 November 2010
Available online 28 December 2010
Keywords:
Spray
Combustion
Visualization
PIV (particle image velocimetry)
PDPA (phase Doppler particle anemometry)
a b s t r a c t
The number of studies on spray and combustion characteristics in IC (internal combustion) engines using
optical techniques has rapidly increased in the past few years due to an increase in the number of
alternative fuels and stricter emission standard regulations. This study investigates better ways of
controlling the combustion process, thus ensuring optimum performance and minimum emission levels
produced during the combustion process. Alternative fuels such as CNG (compressed natural gas), bio-
diesel and hydrogen have been studied by many researchers due to their relative low environmental
impact. Meanwhile, for conventional fuels such as gasoline and diesel, studies have focused on spray
characteristics to provide better air/fuel mixtures in order to produce a cleaner combustion process. Few
experimental works have investigated the effects of modications to the injector itself, for example,
varying the injection rate, injection pressure, etc. In order to provide a better understanding of spray and
combustion characteristics, researchers have studied macroscopic and microscopic parameters using
optical techniques. This paper presents the signicance of spray and combustion study with optical
techniques that have been reported by previous researchers in order to provide in depth knowledge as
assistance to readers interested in this research area.
2010 Elsevier Ltd. All rights reserved.
1. Signicant of study on spray and combustion
characteristics of IC engine
The fast depletion of fossil fuel resources and their contribution
to environmental pollution from ICE (internal combustion engine)
are the major issues that led to increasing demand for efcient and
eco-friendly energy management schemes to be implemented in
industrial, commercial and domestic sectors. One of the feasible
ways to tackle the aforesaid problems is to study the macroscopic
and microscopic parameters of the fuel spray to ensure high engine
performance with low emissions level.
The increasing number of potential alternative fuels also has led
to an increase in interest amongst researchers to study spray and
combustion characteristics of fuels in engine cylinders using optical
methods. Alternative fuels such as biomass, an environmentally
clean and renewable energy source, can be mixed with diesel fuel in
ICEs [1]. This has been studied by some researchers using optical
measuring techniques. Other alternative fuels such as natural gas,
hydrogen and oxygenated fuel have also been studied.
For conventional fuels such as gasoline and diesel, emission
levels can be reduced by enhancing the spray characteristics and
the quality of the air/fuel mixture [2]. It is essential to ensure that
all of the injected fuel has maximum contact with the available air
to produce complete combustion [3]. Few works have investigated
methods of reducing the emissions in ICEs while simultaneously
improving the combustion efciency of the engine. The main
techniques that have been used to reduce NO
x
and PM (particulate
matter) emissions are pulsed injection [4], pilot injection [5,6],
HCCI (homogeneous charge compression-ignition) engines [7],
control of the mixing process and modication of the injection
parameters.
Inorder to understandthe fundamentals of sprayand combustion
characteristics in ICEs, methods such as visualization, PIV (particle
image velocimetry), LIF (laser induceduorescence) andPDPA(phase
Doppler particle anemometry) have been used as diagnostic tools for
in-cylinder measurements. These techniques will be reviewed later
in this paper. Some measurements have been carried out in modied
engines with optical access, and some have been performed in high
pressure and temperature CV (constant-volume) vessels. Studies of
the fuel spray behavior can be classied into two categories:
macroscopic and microscopic. The macroscopic parameters such as
the cone angle and tip penetration can be determined through direct
* Corresponding author. Tel.: 60 45996301; fax: 60 5946025.
E-mail address: mezainal@yahoo.com (Z.A. Zainal).
Contents lists available at ScienceDirect
Energy
j ournal homepage: www. el sevi er. com/ l ocat e/ energy
0360-5442/$ e see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.energy.2010.11.022
Energy 36 (2011) 724e741
visualization methods [8]. Meanwhile, the microscopic parameters
such as velocity, scalar eld and droplet size can be measured using
either PIV or PDPA techniques. It is well known that the efciency of
combustion relies heavily on the quality of the air/fuel mixture,
which is dependent on the fuel injection into the combustion
chamber. Ideally, all of the injected fuel should be in contact with the
maximumamount of available oxygen, toensure that the combustion
of the hydrocarbons can be as complete as possible [9]. In the present
study, a comprehensive survey was carried out in the area of spray
and combustion using optical measuring technique and presented
from the history up to the current stage.
1.1. History
Since 1970s, many researchers focused on the spray and
combustion characteristics of various types of fuel. Few review
papers have been published in the late 70s and in the early 80s on
the fuel droplet size measurement techniques [10,11]. A review by
Chigier [12] gives an outline of atomization and burning of liquid
fuel sprays.
In 1980s, the technique that was used in droplet size measure-
ment can be classied into three classes which are mechanical,
electrical and optical methods. One of the earliest works that was
published on the ame propagation through an air-fuel spray
mixture with transient droplet vaporization in a closed combustor
was reported by Seth et al. [13] in 1980. While in 1981, Banhawy
and Whitelaw [14] had experimentally studied the interaction
between fuel spray and surrounding combustion air at different
degree of air swirl. In 1982, laser tomographic light scattering
technique was used by Yule et al. [15] to study the turbulent
kerosene sprays under cold and vaporization in hot surrounding air
streams. The vapor/liquid visualization in fuel sprays was studied
by Melton and Verdieck using exciplex visualization system in
1985. While in 1991, one of the earliest studies on fuel spray using
laser Doppler systemwas done by Nakabe et al. [16] in studying the
detailed ame structure of liquid fuel sprays. The phase/Doppler
technique was also used by Presser et al. [17] to study the fuel
property effects on the structure of spray ames. While in 1994,
Senda et al. [18] had used visualization and exciplex uorescence
technique to study the fuel vapor concentration in diesel spray.
Over the last 40 years, the techniques have evolved to become
more sophisticated and detailed with important information
obtained from the fuel spray that makes optimization of the spray
parameters possible to researchers.
1.2. Macroscopic spray parameters
The typical spray structure of a DI (direct-injection) fuel spray is
shown in Fig. 1 [19], where the fuel is introduced into the engine
cylinder through a nozzle. As the liquid jet leaves the nozzle, it
becomes turbulent and the outer surface of the jet breaks up into
droplets. In turbulent mixing processes, the turbulent eddies
formed in the shear layer will engulf the surrounding uid in the jet
core, and mixing subsequently takes place on the molecular level at
the two-uid interface [20], thus making the transport phenom-
enon of the uid motion the dominant process in turbulent mixing.
Study of the turbulent transport of the jet uid will directly
determine the distribution of the jet uid concentration and its
vortex motion [21]. As one move away from the nozzle, the mass of
air within the spray increases, the spray diverges, its width
increases, and the velocity decreases. The fuel droplets evaporate as
this air-entrainment process proceeds. The tip of the spray pene-
trates further as the injection proceeds at a decreasing rate.
The region of spray tip penetration becomes longer and wider as
the ambient air temperature increases, due to the lower in-cylinder
gas density at higher temperatures [22]. The longer tip penetration
and larger cone angle are desirable for maximumair utilization and
better combustion due to the increased spray area. In some engine
Nomenclature
S spray tip penetration (m)
Dp pressure drop across the nozzle (Pa)
t time (s)
d
n
nozzle diameter (m)
r
g
gas density (kg/m
3
)
r
l
liquid density (kg/m
3
)
T
g
gas temperature (K)
U
s
(t) instantaneous spray tip velocity (m/s)
D
SM
Sauter mean diameter (m)
D
d
droplet diameter (m)
dn number of droplets
dQ
ch
=dq apparent rate of chemical energy or heat release (J/

)
q crank angle (

)
g ratio of specic heat at constant pressure and volume
p in-cylinder pressure (Pa)
V cylinder volume (m
3
)
T gas mean temperature (

C)
T
w
mean wall temperature
dQ
ht
=dq rate of heat transfer out of the system (W)
l distance (m)
P
u
uorescence signal
h
c
detection efciency of the collection system
U solid angle of collection optics (sr)
V
c
collection volume (m
3
)
f
l
fractional population of the lower electronic state
x
m
mole fraction of the absorbing species
N total number density of the gas molecule (m
3
)
I
v0
spectral intensity of incident laser light (Wm
2
Hz
1
)
A
21
Einstein coefcient for spontaneous emissions
Q
21
collisional transfer coefcient, encompassing all forms
of collisions
B
12
Einstein coefcient for stimulated absorption
I
0
low laser intensity
I
sat
saturation laser intensity
Fig. 1. The denition of macroscopic spray characteristics by Suh et al. [19].
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 725
designs in which the walls are hot and high air swirling is present,
fuel impingement on the walls is desirable. However, in multi-spray
DI diesel combustion systems, with little or no air swirling, over-
penetration results in impingement of liquid fuel on cool surfaces,
which lowers the mixing rates and increases the emissions of the
unburned and partially burned species. The spray tip penetration
can be calculated by the formula developed by Dent [23] as shown
in equation (1), below;
S 3:07
_
Dp
r
g
_
1=4
td
n

1=2
_
294
T
g
_
1=4
(1)
Further detailed studies have examined the spray tip location as
a function of time, and Hiroyasu and Arai [24] correlated their data
for spray tip penetration, S as below;
t< t
break
S 0:39
_
2Dp
r
l
_
1=2
t (2)
t> t
break
S 2:95
_
Dp
r
g
_
1=4
d
n
t
1=2
(3)
where t
break

29r
l
d
n
_
r
g
Dp
_
1=2
(4)
The instantaneous spray tip velocity can be calculated as shown
in equation (5), below;
U
s
t
1
2
S
t
(5)
For jets in the atomization regime, the spray cone angle q was
found using equation (6).
tan
q
2

1
A
4p
_
r
g
r
l
_
1=2

3
p
6
(6)
Finally, using the equations (1) and (6), the projected spray area
and volume of the fuel jet can be calculated.
The injection rates of the fuels can be measured and analyzed
using an injection rate measuring system based on the Bosch
method [25], in which the pressure variation in a tube is monitored
as the fuel is injected into the tube.
1.3. Microscopic spray parameters
The atomization regime is determined by the diesel engine
injection conditions, where the fuel jet is generally in the form of
a cone spray and produces much smaller droplets in comparison
with the nozzle exit diameter. However, at low jet velocities, the
droplet size is larger than the nozzle diameter, due to the unstable
growth of surface waves caused by surface tension. The further
increment in jet velocity leads to a breakup in the atomization
regime of the jet outer surface at or before the nozzle exit plane,
creating droplets with an average diameter that is much smaller
than the nozzle diameter. Atomization of the liquid fuel into a large
number of small droplets is required to create a large surface area
for the fuel to evaporate.
During the injection period, the droplet size distribution at
a given location may vary not only with time but also with injection
conditions, such as the difference in droplet size at the spray core
and spray edge, trajectories of individual droplets and the initial
velocity due to the atomization process. In order to characterize the
spray, expressions for droplet size distribution and mean diameter
are required; the latter can be represented by the Sauter mean
diameter [26] as shown in equation (7).
D
SM

__
D
3
d
dn
____
D
2
d
dn
_
(7)
After the atomization process, spray evaporation of the jet fuel is
essential before it can be mixed with air. There are three
phenomena involved in the droplet evaporation process: deceler-
ation of the droplet due to aerodynamic drag, heat transfer to the
droplet from the air and mass transfer away from the droplet by
vaporization. As the droplet comes into contact with hot air inside
the combustion chamber, the fuel vapor pressure and evaporation
rate will increase due to heat transfer fromthe air. The vaporization
process of the spray is very important as it will affect the ignition
stability and ame propagation.
Few studies have been conducted investigating the structure of
non-evaporative and evaporative sprays. In non-evaporative tran-
sient sprays, the inner structure is regularly spaced and branched
[27]. The spacing between the branches is in the order of a few
millimeters, and it is nearly constant throughout the injection
period. Meanwhile, Kosaka and Kamimoto [28] investigated the
spray structure of an evaporating spray using SSI (silicone particle
scattering imaging). The remaining silicone oil droplets in the
evaporating spray represent the fuel vapor concentration. It was
found that the silicone oil remained un-evaporated at a tempera-
ture of 640 K and that the evaporating spray at higher temperatures
has a structure that differs from that of non-evaporating sprays.
Yuyama et al. [29] developed a novel entropy analysis method
based onstatistical thermodynamics. Entropyanalysis candene the
degree of homogeneity in the fuel concentration. The high concen-
tration of the liquid-phase distribution tip occurs at the wall of the
combustion chamber before evaporation, thus leading to a high
possibility of emissions of unburned hydrocarbon. Therefore, a high
degree of homogeneity is desired in engine combustion, as it will
produce more complete combustion and reduce engine emissions.
1.4. Combustion characteristics
The fundamental step of the combustion process in SI (spark-
ignition) engines begins with the fuel and air mixing and being
induced into the combustion chamber through the intake valve. At
the end of the compression stroke, the combustion is initiated by an
electric discharge at the spark plug. The turbulent ame develops
and propagates inside the combustion chamber and is then extin-
guished at the end of the power stroke.
Meanwhile, the difference between CI (compression-ignition)
engines and SI engines lies in their compression ratios, where the
compression ratio of CI engines is much higher, in the range of
12e24, to allowfor self-ignition of the diesel fuel. The valve timings
used in CI engines are similar to those of SI engines. The air at close-
to-atmospheric pressure is induced during the intake stroke and
then compressed to a pressure of about 4 MPa and a temperature of
about 800 K during the compression stroke. Fuel injection into the
engine cylinder starts at about 20

before the TC (top center).

The liquid fuel jet atomizes into droplets and evaporates inside
the engine combustion chamber and then mixes with air within the
combustible limit. Within a short period, the air temperature and
pressure rise above the fuels ignition point. After a short delay
period, spontaneous ignition of the fuel/air mixture initiates the
combustion process, and the cylinder pressure rises above the non-
ring engine level. The ame spreads rapidly through the portion of
injected fuel that has already mixed with sufcient air for burning.
As the expansion process proceedss, mixing between fuel, air and
burning gases continues, accompanied by further combustion.
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 726
An important engine combustion diagnostic measurement is
the in-cylinder pressure trace. Using this pressure measurement,
considerable information about the nature of the combustion
process under any operating condition can be collected. A typical
in-cylinder pressure diagram for SI and CI engines is shown in
Fig. 2, from Zainal and Soid [30]. Using the in-cylinder diagram, the
most important information can be obtained by performing a so-
called heat release analysis. This analysis will reveal the rate of
mass burned and the volume fraction engulfed by the ame as
functions of the crank angle. For SI engines, the chemical energy or
gross heat release rate can be calculated using equation (8).
dQ
ch
dq

g
g 1
p
dV
dq

1
g 1
V
dp
dq
V
cr
_
T
0
T
w

T
T
w
g 1

1
bT
w
ln
_
g 1
g
0
1
_
dp
dq

dQ
ht
dq
_
(8)
For CI engine operation, the heat release rate [31] is given by
equation (9). A typical heat release diagram for CI engine operation
is shown in Fig. 3.
dQ
n
dt

g
g 1
p
dV
dt

1
g 1
V
dp
dt
(9)
In diesel engines, the ignition delay is dened as the time
interval between the SOI (start of injection) and the SOC (start of
combustion). The ignition delay occurs because both physical and
chemical processes must take place before a signicant fraction of
chemical energy of the injected fuel is released. The start of the fuel
injection is usually determined by a needle-lift indicator which
traces the movement of the injector needle with respect to crank
angle. The SOC can be determined in several ways such as the point
where ring pressure exceeds motoring pressure or where heat
release begins or maybe when the rst visible light from combus-
tion is observed. The duration of this ignition delay period is
especially important in diesel combustion since it affects both the
performance and emissions. The net heat release rate during this
period usually turns out to be negative due to the heat loses to the
cylinder walls and the fuel vaporization.
In premixed or rapid combustion phase, the fuel and air that are
mixed and prepared during the ignition delay period burn rapidly,
leading to high heat release rates and rapid cylinder pressure rise.
Since the amount of fuel and air are mixed within the ammability
limits depends on the duration of the ignition delay period, the
longer the ignition delay, the more prepared fuel-air mixture with
higher initial heat release rates and pressure rise. Thus in order to
avoid excessive heat release rates and to keep rates of pressure rise
acceptable, the ignition delay period should be as short as possible.
After the premixed fuel-air mixture has been burned, the rate
of combustion in mixing-controlled combustion phase is controlled
by the rate at which fuel-air mixture becomes available for burning
which is largely dependent on the fuel injection rate and fuel vapor-
air mixing rates. The heat release in this phase is typically of a lower
magnitude and occurs over a much longer duration compared to
the rapid combustion phase. This is because the fuel-air mixing
time scales are much higher than the chemical reaction time scales.
As the injected fuel continues to burn, the piston expands and the
heat release rate decreases.
Late combustion phase is characterized by very low heat release
rates due to the combustion of the remaining fuel, COand soot. This
phase is usually terminated by the temperature drop during
expansion which slows down the kinetics of the nal burning.
1.5. Conventional techniques
Previously, the technique that was used in droplet size measure-
ment can be classied into three classes which are mechanical
method (droplet capture, cascade impaction, frozen drop and
wax methods and sedimentation technique), electrical methods
(WickseDukler technique, the charged wire probe and the hot wire
anemometer) and optical method (photography, holography, laser
diffraction, laser anemometry and various other techniques based on
light scattering). However there is not muchinformationobtainedfor
the conventional technique, except for the large scale model.
Palopski [32] had used a large scale model to study the spray
characteristics of fuel such as the droplet size and velocity. The
nozzle size used was larger than the normal size causing a decrease
in the droplet velocity and increase in droplet size. Walzel [33] had
also used large scale models in spray formation studies and small
scale models for black liquor nozzles of recovery boilers and heavy
oil nozzles of blast furnaces that was used in splash plate nozzle
studies. More work using large scale models have been used by
Soteriou et al. [34] to study the fuel ow inside the diesel nozzle.
However, the used of large scale model only helped in under-
standing the break up and atomization mechanism but might be
inaccurate for diesel spray characteristics as proven by Rantanen
et al. [35].
2. Optical techniques
Optical techniques are commonly used for macroscopic and
microscopic fuel spray and combustion characterization. However,
these techniques can often be difcult to implement and very
expensive. According to Roisman et al. [36] the spray, especially in
the neighborhood of the nozzle, is very dense; making the use of
optical measurement techniques very difcult. Some examples of
Fig. 2. In-cylinder pressure comparison between experiment and modeling at
2400 rpm and 30% of brake power by Zainal and Soid [30].
Fig. 3. Phases in Diesel Engine Combustion Process by Heywood [26].
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 727
optical techniques are visualization, PIV, LIF and PDPA. The visual-
ization technique has been reviewed by Chigier [37], while the
optical measurement techniques such as LIF have been reviewed
previously by Zhao and Ladommatos [38].
2.1. Visualization technique
The most frequently used optical measurement technique for
macroscopic characteristic of fuel spray measurement is the direct
visualization method. The visualization of digital imaging either in
solid state or continuous recording can be employed to expose the
conditions of the spray and the combustion process. The macro-
scopic parameter such as spray geometry, penetration depth, wall
impingement, and subsequent combustion can be exposed using
this technique. For the purpose of simulating the actual spray
and the combustion process in IC engines, a high pressure and
temperature CV vessel can be used as the combustion chamber,
and shadowgraph photos of the fuel spray and the combustion
process can be acquired using a high-speed digital camera. The
typical experimental setup for visualization technique was shown
in Fig. 4.
2.1.1. Spray and combustion characterization using visualization
2.1.1.1. Spray and combustion characteristics. Diesel spray injection
and combustion was studied by Bougie et al. [39] in a high pressure
and temperature cell. As a novelty, they had observed quasi-
simultaneous imaging of spray propagation and combustion using
two ICCD (intensied charge-couple device) cameras with double
framing options. Two dimensional ash light shadowgraph, laser
elastic scattering and chemiluminescence were used to investigate
the fuel propagation, and laser-induced incandescence was used to
study the soot formation. The analysis shows the fuel vapor cloud
extending well beyond the liquid core and close to the nozzle tip.
The ignition also takes place close to the spray tip and soot is
formed in the vapor core region at the tip. The characteristics of an
evaporating diesel spray and the ame distribution in the
combustion chamber of a DI diesel engine also was investigated by
Yoshizaki et al. [40].
The macroscopic diesel fuel spray was studied by Klein-Douwel
et al. [41] using high-speed digital imaging in a high pressure cell to
provide a better and more fundamental understanding of the
interaction between the fuel injection and the spray propagation. In
this work, they had presented the experiment and the algorithmfor
geometrical analysis and explain the way to determine the local
spray angle, cone angle and spray penetration.
The ame structures of wall-impinging diesel sprays injected by
group-hole nozzles were investigated by Gao et al. [42] in a CV
combustion vessel, as shown in Fig. 5. It was found that an asym-
metric ame structure emerged in wall-impinging group-hole
nozzle sprays. Group-hole nozzles, with a large included angle,
yield a higher overall OH chemiluminescence intensity. The wider
the ame, the greater the proportion of high OH intensity, thus
providing better fuel/air mixing and improving the combustion
process.
Kostas et al. [43] experimentally studied the initial stages of
diesel fuel spray development in a CV vessel. The fuel spray was
observed for a duration of 0.5 ms after the SOI using an ultra-high
speed camera. It was found that the tip penetration shows linear
dependencies at the initial stages of spray development. Moreover,
the spray tip velocity was found not to be constant, but propor-
tional to

t
p
.
Nishida et al. [44] investigated the mixture formation process of
sprays injected from a two-hole nozzle for DISI (direct-injection
spark-ignition) engines, aimed to achieve better air utilization,
using the LAS (laser absorption-scattering) technique. Their results
showed that, as the HAA (hole-axis angle) of the nozzle decreased,
the droplet coalescence increased and the vapor mass decreased.
The tip penetration also increased with decreasing HAA. Addi-
tionally, an increased injection pressure signicantly increases the
vaporization rate and the tip penetration rate.
The characteristics of spontaneous ignition and combustion
with unsteady high-speed gases such as natural gas and hydrogen
fuel jets were investigated by Ishiyama et al. [45] using visualiza-
tion methods in a CV vessel as shown in Fig. 6. The fundamental
data of engine combustion that was fueled by natural gas and
hydrogen were observed in order to determine a method for
controlling the combustion. It was found that a higher temperature
is needed for natural gas ignition, and for hydrogen fuel, the igni-
tion delay is shorter, as compared to natural gas.
The spray and combustion characteristics of DME (dimethyl
ether) were experimentally studied by Wakai et al. [46] using a CV
vessel at 3.1 MPa and 823 K. It was found that the DME spray was
ignited 2.6 ms after the SOI, with the rst ignition occurring around
the middle portion of the spray and the ame spreading in the
downstream direction. Further observation showed that the ame
was non-luminous and that light emission fromthe ame was very
weak. Under lower ambient pressures (2.1 and 1.1 MPa), the igni-
tion delay of the DME spray was shorter than that of a diesel fuel
spray, whereas it was almost the same as the diesel fuel delay at
higher ambient pressures (4.1 and 3.1 MPa). Moreover, at an
ambient pressure of 1.1 MPa, the DME spray could be ignited at
temperatures >770 K, in contrast to the diesel fuel, which could not
be ignited.
Bang and Lee [47] had studied the fuel injection characteristics
and spray behavior of biodiesel blended with methyl ester using
visualization technique. The experiments were conducted at
various blending ratio; B25, B50 and B75 fuels. From the analysis it
was found that a shorter injection delay, shorter injection duration
and higher peak value of injection rate was noted by lower blending
ratio. The spray penetration analysis also shows the blended fuel
has a similar spray length at the same condition except for the neat
DME fuel. It was also found that the DME evaporated more rapidly,
but a wider cone angle was observed when biodiesel and blended
fuels were tested.
The spray properties of different blends of ethanol-gasoline
(25%, 50%, 75% and 100% ethanol) was studied by Gao et al. [48]
under various ambient conditions using high-speed schlieren Fig. 4. Visualization setup used by Bougie et al. [39].
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 728
photography technique. Fromthe analysis, the similar spray pattern
was observed for both blended fuel and pure gasoline. At low
ambient pressure, there is better evaporation with the increase of
ethanol fraction, indicated by the decrease of tip penetration and
increase in spray angle. However, at elevated pressure, there is no
difference in spray penetration among the blends. Moreover, the
spray angle for all blends was kept almost constant in the fully
developed stage.
The inuences of different injection strategies and biodiesel
blends on the spray and combustion development were studied by
Fang et al. [49] in an optical HSDI (high-speed direct-injection)
engine. In order to investigate the effect of different injection
strategies, three injection timings were chosen to the four diesel/
biodiesel blends that are pure European low sulfur diesel (B0), 20%
(by volume) soybean biodiesel mixed with 80% low sulfur diesel
(B20), 50% soybean biodiesel mixed with 50% low sulfur diesel
(B50), and pure soybean biodiesel (B100). It was found that the heat
release rate of biodiesel is dominated by premixed combustion and
that the heat release peak decreases with a slower injection timing.
Combustion analysis also showed that the ignition and heat release
rate peak occur later with increasing biodiesel content. Visualiza-
tion of the combustion process also showed that stronger fuel
impingement and a longer tip penetration were produced with
increasing biodiesel content. An extension work on biodiesel was
done by Fang et al. [50] on the effects of advanced injection strat-
egies and NO
x
emissions using the same optical HSDI engine.
Desantes et al. [51] had studied the macroscopic injection
parameters of three biodiesel blends on DI diesel-injection using
a standard injection system derived from a 4-Stroke diesel engine.
The experiments were carried out over a commercial diesel fuel,
with 5.75% of RME (rape methyl ester) categorized as B5, another
with 30% of RME (B30), and a pure RME fuel. The analysis was done
on the injection rate shape, spray force, spray tip penetration and
cone angle in non-evaporative conditions. It was found that the fuel
property such as density of the fuel highly affects the injection
parameters. Higher viscosity of the RME makes the injector needle
lift slower thus resulted in longer penetration, narrower cone angle
and smaller spray volume.
An extension work on RME fuel was done by Pastor et al. [52] to
study the liquid-phase fuel penetration, soot luminosity measure-
ments and time-resolved intensied imaging to detect apparition of
CHand OHchemiluminescence and early soot radiations. The liquid-
phase fuel penetration and soot luminosity were measured using
high-speed imaging, and the chemiluminescences of both CH and
OHhave been observed using intensied imaging. Fromthe analysis,
it was found that the fuel penetration increases linearly with
blending rates (RMElongest, 30%of RMEmediumand5%of RMEwas
the shortest). The chemiluminescence of CH found led to a decrease
of ignition delay by about 50 ms for an increase of RME blend rate.
Moreover, the high-speed imaging of OHchemiluminescence shows
that the RME rate did not show any signicant inuence, as this
parameter is controlled by physics rather than by chemistry.
There were two studies done on the combustion near the valves
regions. The fuel deposition burning in PFI (ported fuel injection) in
SI engine was studied by Merola and Vaglieco [53] to investigate
the diffusion-controlled ame to evaluate the ame propagation
and the chemical species involved during the combustion. Morela
et al. [54] had also studied on the effect of the fuel injection strategy
using the same rig. Single and double fuel strategies with two
injection pressures were investigated to study the diffusion-
controlled ames in valves region and on the cylinder wall.
2.1.1.2. Effect of nozzle parameters. Payri et al. [55] investigated the
effect of cavitation on the internal ow and macroscopic spray
characteristics in a diesel-injection nozzle using visualization
methods. In this work, two bi-orice nozzles were used, which were
cylindrical and convergent (conical). In order to avoid obscuring the
windows, nozzles with two orices were used and the injector is
positioned so that the sprays are situated perpendicular to the Fig. 6. Constant Volume Vessel used by Ishiyama et al. [45].
Fig. 5. Schematic of the constant-volume combustion vessel and optical setup for direct ame and OH chemiluminescence imaging used by Gao et al. [42].
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 729
camera and ash window. This also allows a characterization of each
spray to be made without interferences from others. It was found
that cavitation leads to an increment of the spray cone angle. Based
on internal ow analysis, when cavitation is initiated, the whole
outlet velocity increases due to a reduction in the cross section of
the liquid phase in the outlet section of the hole. The structure and
evolution of cavitation and its inuence on jet patterns were
previously studied by Ganippa et al. [56]. More work on the effect of
nozzle cavitations of diesel and biodiesel fuels was done by Suh et al.
[57] and they had found that the cavitation of biodiesel fuel would
occur at a higher cavitation number.
The impact of injection pressure was investigated by Delacourt
et al. [58] using digital imaging techniques. As pressure is one of the
most important parameters, the spray characteristics at very high
pressures (up to 250 MPa) were evaluated. For the experimental
setup, images of the spray development were captured using a fast
video camera under direct lighting. In this work, spray character-
istics such as cone angle, spray area and tip penetration were
observed (Fig. 7). The impact of ultra-high injection pressure (up to
300 MPa) of biodiesels on spray characteristics was also investi-
gated by Wang et al. [59].
The effects of orice diameter on a diesel fuel jet ame were
investigated by Pickett and Siebers [60] in a CV combustion vessel.
The orice diameters were varied from45 mmto 180 mm, and it was
found that, with decreasing orice diameter, the lift-off length (the
farthest upstream location of high-temperature combustion)
decreases, but the estimated amount of air entrained increases,
thus resulting in an overall average equivalence ratio decrease in
the initial combustion zone. The total soot incandescence signi-
cantly decreases as the orice diameter decreases, indicating that
a smaller amount of soot is formed relative to the amount of fuel
injected. Moreover, the normalized ame length appears to
decrease as the orice diameter increases.
The effect of the diesel nozzle geometry on spray liquid-phase
fuel penetration in evaporative conditions was investigated by
Payri et al. [61] using visualization methods. Three six-hole sac
nozzles used differ in the orice degree of conicities and were
characterized geometrically and hydraulically were used. The
liquid-phase fuel penetration in an evaporative environment was
also measured. It was found that the velocity and mass ow rate
were strongly inuenced by cavitation, but the momentum ux
was not inuenced by the appearance or development of cavitation.
Based on a LL (liquid length) analysis, in general, for a given
temperature, a shorter LL means faster and better air/fuel mixing.
An extension work on the effect of nozzle geometry was done by
Payri et al. [62] and focused on the CH and OH chemiluminescence
techniques to examine the combustion process.
Meanwhile, the effects of the injection angle on the combustion
processes of HSDI diesel engines were studied by Fang et al. [63]
using an optical engine. In this work, two different injectors, with
70-degree and 150-degree tips were used. It was found that the
heat release patterns were different for both injectors. Fuel wall
impingement was found on the 70-degree tip on the rst injection.
The 150-degree tip produced a weak ame for the rst injection,
but the 70-degree tip produced two types of ames: an early weak
ame and a late luminous lm combustion ame. The spray
patterns for 70-degree and 150-degree tips are shown in Fig. 8.
2.1.1.3. Fuel injection system. A swirling injector was also studied
by Moon et al. [64] using visualization methods. In this work, they
had used an extended tiny probe (0.6 mm outer diameter and an
inner diameter of 0.4 mm) to measure the static pressure inside the
spray and along the nozzle axis. The visualization of spray shows
that the extended probe can access up to 0.5 mm from the nozzle
without disturbing the spray shape. In order to obtain the macro-
scopic spray development under different operating conditions,
Mie-scattered spray images were obtained using a laser sheet
formed by a 6 Wcw Ar-ion laser, a cylindrical lens with focal length
6.5 mm, and a plano-convex lens with a focal length of 1000 mm. A
CCD (charge coupled device) camera (PCO Inc. Sensicam) was
utilized with a resolution of 1280 1024 pixels with an exposure
time of 10 ms. It was found that the air pressure drop was greater at
higher injection pressures. The air pressure at the nozzle exit also
showed a higher value at higher fuel temperatures. Moreover, the
pressure drop at high fuel temperatures was greater than that of
atmospheric temperature. Further downstream, the reduced air
pressure returned to atmospheric pressure.
2.1.1.4. Cyclic variations in IC engine. Wang et al. [65] studied the
cyclic variations of DI combustion of natural gas hydrogen blends
using visualization methods in a CV vessel. Flame propagation
photographs and pressure data were acquired at various hydrogen
volumetric fractions in the fuel blend. In order to study the effect of
cyclic variations, the authors repeated 50 combustion experiments
under the same initial condition for various hydrogen fraction fuel
blends and equivalence ratios of 0.6 and 0.8. The amount of data can
provide the analysis of cyclic variations using CVvessel. It was found
that cyclic variations were initiated at the early stages of ame
development. The addition of hydrogen caused a ame kernel that
was nearly concentric with the spark electrode. Moreover, an
increase in the hydrogen fraction decreased the cyclic variation.
2.1.1.5. Engine emission. Kitamura et al. [66] studied NO
x
control in
DI PCCI combustion with a CV vessel using n-heptane that was
injected into the combustion chamber using a single-hole nozzle. In
order to obtain the NO
x
(NONO
2
) concentration, the collected gas
Fig. 7. Spray angle calculation q (the negative of the image) by Delacourt et al. [58].
Fig. 8. Spray in quiescent air for the 150-degree (left) and 70-degree (right) injector
tips by Fang et al. [63].
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 730
was then analyzed using a chemiluminescence NO
x
analyzer
(YANACO ECL-88A). It was found that enhanced fuel/air mixing
reduced NO
x
accompanied by a delayed hot ame, and under low
oxygen mole fraction conditions, a lower heat release rate was
obtained. Based on the analysis of a xed nozzle orice diameter,
the nal NO
x
emission was reduced with increasing injection
pressure. It was concluded that the enhancement of fuel/air mixing
reduces the nal NO
x
emissions, but when the quantity of the
injected fuel exceeds a certain level, the injection conditions have
only a slight effect on the nal NO
x
emissions. The shadowgraph
images and pressure history are shown in Fig. 9.
2.2. Particle image velocimetry (PIV) technique
The experimental arrangements for PIV measurements are
shown in Fig. 10. PIV is a measurement technique for obtaining the
instantaneous whole eld velocities. The PIV will measure the
distance traveled by particles in the ow within a known time
interval. These particles are known as seeding particles. The seed-
ing particles are added to follow the ow, and in order to detect
their movement, an area of the ow is illuminated by a light sheet
that can be generated by a double pulses laser and optical system.
To detect the position of the illuminated seeding particles, a CCD
camera is located at right angles to the light sheet. The particle
position will appear as light specks on a dark background on each
camera frame. The pulsing light sheet and camera are synchronized
so that the particle position at the light pulse number one is
obtained at frame number one and particle position from pulse
number two is on frame number two.
The images are divided into rectangular regions, known as
interrogation areas. The interrogation areas from each frame are
correlated to produce an average particle displacement vector to
produce a vector map of average particle displacements and the
displacement vectors are converted into a map of so-called raw
velocity vectors. More information about PIV can be found in
a report by Kompenhans and Khler [68].
2.2.1. Spray characterization using PIV
2.2.1.1. Turbulent premixed ame. Stella et al. [69] had studied
three-component PIV (Fig. 11) measurements in a turbulent pre-
mixed ame. As the conventional PIV allows the measurement of
the in-plane components of the velocity, the stereoscopic PIV
conguration also yields the out of plane components. In this
work, stereoscopic PIV technique was used to investigate the
turbulence jet ame. Using the stereoscopic PIV, the availability of
planar three components velocity elds signicantly enhanced the
capability of inferring three-dimensional ows in a correct and
detailed manner. Three-component planar velocity elds (2D3C)
were measured using a stereoscopic PIV system. Two recording
cameras were used to observe the motion of seeding particles
from different directions. From the third component, these two
cameras recorded different apparent particle displacements. The
premixed jet ames were produced with a laboratory burner and
the mixing between the fuel (propane) and air was injected from
four orthogonal orices. From the analysis, it was found that the
stereoscopic PIV technique was capable of measuring three-
component planar velocity elds that associated with topological
information on the instantaneous reacting surface, deduced by the
distribution of seeding particle density in the recording image.
This technique is very effective for turbulent combustion process
investigation.
2.2.1.2. Spray behavior and distribution. Lee et al. [67] experimen-
tally studied the spray behavior and fuel distribution of GDI
(gasoline direct-injection) injectors using entropy analysis and PIV
methods. Their results showed that the fuel spray propagates more
widely in the downstream direction. The entropies of the whole
area of the spray are higher when the ambient temperature is
higher than the fuel evaporation temperature. The spray of the
large injector attached angle is distributed at the center of the
cylinder, as compared to the small injector attached angle. More-
over, the droplets are concentrated near the injector tip at the
beginning of the injection, and they propagate intermittently, with
the mixture nally achieving a homogeneous distribution.
The spray structures of DMM (dimethoxy methane), DMC
(dimethyl carbonate ) and DME were investigated by Wu et al. [70]
using laser-based visualization and PIV. Their analysis showed that
the spray of oxygenated fuel shows an umbrella-shape structure,
a larger spray angle and a shorter spray tip penetration, as
compared to diesel fuel. Moreover, it also presents a weak large-
scale heterogeneity and branch-like structure, ner droplets,
a stronger interface and more violent vertical motion. Furthermore,
their analysis also showed that the viscosity of oxygenated fuel has
a signicant effect on the atomization behavior.
An investigation of the characterization of the GDI oweld and
stratication effects of fuel spray in a visualization engine using
DPIV (digital particle image velocimetry) and entropy analysis were
conducted by Lee and Lee [71]. Their analysis revealed a relation-
ship between vorticity and entropy, which plays an important role
in the diffusion process. The homogeneous diffusion rate of the
spray was quantied using entropy analysis, and with this method,
it was found that a homogeneous mixture distribution is more
effective than the spray concentration in dissipating the
momentum of the surrounding air in order to alter the injection
timing. Moreover, an increase in the injection timing increases the
homogeneous diffusion rate.
Fig. 9. Shadowgraph images and pressure history under the simulated diesel condition
(Ambient condition: pi 4 MPa, Ti 900K, rO2 21%, Injection condition:
dN 0.18 mm, pinj 120 MPa, mf 19.5 mg) by Kitamura et al. [66].
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 731
Myong et al. [72] studied the spray inner structure of an evap-
orating multi-component fuel using the PIV method in a CV vessel,
as shown in Fig. 12. The tested single-component fuels were
i-octane, n-dodecane and n-hexadecane. The multi-component
fuels were designated as A, B and C, as shown in Table 1. Their
analysis revealed that, in a HI-BP (high boiling fuel), the inner
structure was very heterogeneous and a high concentration of the
liquid phase in the spray tip occurred near the combustion chamber
wall before evaporation, thus increasing the possibility of emissions
of unburned HC (hydrocarbon). Moreover, it was found that evap-
oration was promoted by the addition of a LO-BP (low boiling point
fuel), because it will lower the boiling point of the fuel.
2.3. Laser induced uorescence (LIF) technique
LIF uses light emission from an atom or molecule following
excitation by a laser beam. The principle mechanism in LIF is the
excitation of a molecule from its initial lower electronic state to an
upper electronic state due to laser source. The processes in
producing the LIF begin with excitation of a molecule by laser
source, followed by the molecule returning to its initial state by
stimulate emission. Then the molecule excites to higher state due to
absorption of additional photon. Next, the internal energy produces
rotational and vibration energy transfer due to collisions between
the molecules. Later, the interactions between the individual atoms
of the molecule produce internal energy transfer and dissociation.
Finally, the originally and nearby populated state return to lower
states through the emission of light and at the same time producing
the LIF. The uorescence signal can be expressed using equation
(10);
P
flu
h
c
UV
c
f
l
Tx
m
NI
v0
A
21
A
21
Q
21
B
12
1
I0
Isat
(10)
Fig. 10. Schematic diagram of experimental arrangement for PIV used by Lee et al. [67].
Fig. 11. Scheme of PIV stereoscopic conguration according to the angular displace-
ment method (ADM) used by Stella et al. [69].
Fig. 12. Schematic cross-section of the constant-volume vessel used by Myong et al.
[72].
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 732
LIF can be used to visualize the liquid droplet and fuel vapor at
the same time. In fuel spray, although the liquid and vapor may
present at the same time, the signal fromthe liquid phase normally
is much stronger than the signal from fuel vapor. The image can be
obtained using an intensied CCD camera. To obtain the uores-
cence image and to avoid elastically scattered light, an objective
lens that is opaque to UV (ultra violet)-rays can be used with a color
lter attached to the front of the lens. More information on LIF
technique can be found in a review by Zhao and Ladommatos [38]
and Desantes et al. [73] also had studied the limitations of the PLIEF
(planar laser induced exciplex uorescence) in diesel spray. The
typical experimental setup for PLIEF technique is shown in Fig. 13.
2.3.1. Fuel phase, droplet sizing and species concentration in fuel
spray
2.3.1.1. Fuel phases in spray. The spray structure and distribution
characteristics of liquid and vapor phases in non-evaporating and
evaporating GDI fuel sprays were studied by Choi et al. [74] using
LIEF (laser induced exciplex uorescence) technique. The dopants
used were 2% uorobenzene and 9% DEMA (diethyl-methyl-amine)
in 89% solution of hexane by volume. Meanwhile, PDA (phase
doppler anemometry) technique was used to study the internal
structure of the spray. The analysis showed there were two regions
in the spray that was exposed by uorescence intensity of liquid
phase they were cone and mixing regions. In the mixing region,
vortex ow of the vapor phase was observed.
2.3.1.2. Droplet sizing. Droplet combustion of methane in strained
counter-ow diffusion ames was studied by Mercier et al. [75]
using planar LIF. A mixture of CH
4
diluted with N
2
was injected
from the lower nozzle and air from upper nozzle as shown in
Fig. 14, and a monodisperse droplet generator was used to inject
acetone droplet through the ame. PLIF will measure the OHradical
and acetone was used to measure the spatial distribution of the fuel
vapor and the structure of the ame front. The analysis showed that
the droplet crossing strongly distorted the ame front. Further
investigation revealed that, high velocity droplet led to local
extinction of the ame front, and low velocity droplet could ignite
within the ame front and burn on the oxidizer side.
An LIF image processing technique for measuring drop sizes in
a pre-swirl spray was developed by Lee and Nishida [76]. A
microscopic imaging technique was applied to obtain high spatial
resolution LIF tomograms of the pre-swirl spray. 1 mm thick
Nd:YAG laser sheet was used as a light source and the image was
captured by high resolution CCD camera. Image processing was
used to obtain the droplet size. The laser sheet-straddling large-
sized droplets representing low intensity were screened effectively
by applying the mean intensity ratio, and the small-sized droplets
are relieved reasonably through the inspection of the line prole.
They had concluded, the depth of eld effects in droplet sizing by
laser sheet imaging were effectively measured, but there were
uncertainty with the Gaussian energy distribution of the laser sheet
and the non-uniform distribution of the concentration of the
uorescent dye inside the droplets.
2.3.1.3. NO concentration in IC engine. The in-cylinder NO (nitrogen
oxide) concentrations were studied by Verbiezen et al. [77] at
various injection timing in a heavy duty diesel engine. The presence
of the diffusion ame is related to the onset of NO formation within
the laser probe volume observed using images from high-speed
combustion. Fromthe schlieren images, it was found that the initial
combustion is too fuel-rich for thermal NOformation. Further study
using numerical and experimental methods prove that the thermal
mechanism is the major pathway for NO formation.
2.4. Phase Doppler particle anemometer (PDPA) technique
PDPA technique can be used for droplet sizing investigation in
the fuel spray. The experimental measurement arrangement for
PDPA is shown in Fig. 15. The PDPA system consists of an Ar-ion
laser, a transmitter, a receiver and a data acquisition system. To
obtain time-resolved data, the signal analyzer is synchronized with
the injector driver using a digital delay generator. By averaging the
captured droplets at all of the measurement points, the SMD
(sauter mean diameter) at a specic time is obtained.
The atomization characteristic such as droplet velocity distri-
bution is measured using PDPA technique. In this system, the Ar-ion
laser is used as the light source and the signal analyzer is
synchronized with the fuel injection signal to obtain the transient
characteristics of the fuel atomization. The laser output and PMT
(photomultiplier tube) voltage are set based on the data acquisition
rate and signal intensity. When fuel droplets pass through the
Table 1
Experimental condition and fuel composition.
Injection nozzle Type hole nozzle DLL-p
Diameter of hole (dn, mm) 02
Length of hole (Ln, mm) 0.8
Test fuel Single-component fuel
(C
8
H
18
, C
12
H
26
, C
16
H
34
)
Multi-component fuel
(Fuel A, Fuel B, Fuel C)
Ambient temperature
(Tamb, K)
500, 600, 700
Fuel temperature
(Tfuel, K)
300, 362
Ambient density
(qamb, kg/m3)
5, 15
Injection pressure
(Pinj, MPa)
72
Ambient gas Ar
Injection quantity
(Q, mg)
12
Fuel component Mass fraction
of MCFs
i-Octane:n-dodecane:n-hexadecane (C8H18:C12H26:C16H34) Fuel A 6:31
Fuel B 1:1:1
Fuel C 1:3:6
Injection duration (tinj, ms) Fuel A 1.82
Fuel B 1.82
Fuel C 1.81
Fig. 13. Experimental layout for PLIEF technique by Desantes et al. [73].
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 733
measurement volume of fringe created by the laser beam from the
Ar-ion laser, the dispersed beam is detected at the receiver. The
detected phase difference and Doppler signal frequency are then
converted to the mean droplet size and velocity. The fuel droplet
size is obtained from the PDPA system for the analysis of the fuel
atomization. The cutoff range of the droplet diameter measurement
by using the droplet measuring system is set about 2e100 mm, and
more than 20,000 spray droplets are collected and averaged for
each measuring point. The representative SMD at a specic time is
determined by averaging the captured droplets over all of the spray
measurement area.
In order to utilize this technique, interference fringe pattern is
needed to be created within a small volume in the spray. A scat-
tered light signal is generated when a droplet moves through
the small volume. The velocity of the droplet will be related to the
temporal modulation, while the droplet size will be related to the
spatial frequency. However, Zhao and Ladommatos [38] had stated
that PDPA is not suitable for high density sprays; due to it only
needs one droplet in the probe volume during the measurement.
Moreover, this technique also cannot provide size distribution in
a single transient injection. However, important information such
as fuel droplet size and velocity, density, volume ux and time
resolved can be obtained using this technique.
2.4.1. Fuel atomization measurement using PDPA
2.4.1.1. Structure of spray and fuel ow. Rantanen et al. [35] had
studied the measurement of a diesel spray of a normal size nozzle
and a large scale model using laser Doppler anemometer technique.
The aim of this work is to study the effect of scaled experiments to
the spray formation and atomization process. From the analysis of
two test rigs (normal and scale 4:1) it was found that, the spray
angle was 15% higher in the normal scale and the droplet size was
scaled up to 80%. They had proved, scaling might caused inaccuracy
in diesel sprays, but it can still provide valuable information about
the break up and atomization mechanisms.
ChanandTurner [78] hadmeasuredthe velocity inside a motored
ICE using three-component laser Doppler anemometry as shown in
Fig. 16. The aim of their study was to investigate the structure of the
ow inside the combustion chamber of a motored ICE. A single
cylinder 4-stroke research engine driven by an electric motor that
was built for high-speed photographic studies was used. For three-
component LDA optical system, a six-beam LDA system was
assembled by combining a single-component Dantee 55X conven-
tional modular optical system with a two-component four beam
Dantee 60X Fibre-ow optical system. All the six beams were
directed into the cylinder volume through the semi-circular window
incorporated within the cylinder head of the model engine. The
Fig. 14. Droplet combustion of methane in strained counter-ow diffusion ames measurement used by Mercier et al. [75].
Fig. 15. Droplet particle analyzer system using PDPA used by Suh et al. [19].
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 734
experiment of inclined inlet port conguration has revealed that the
inclined port creates ow conditions more favorable to mixing
process inthe cylinder, because it generates a regionwitha relatively
highshear andstrong recirculationzones inthe cylinder. Moreover it
produces more nearly homogeneous owstructure at TDC (top dead
center) during the compression stroke.
Doudou [79] experimentally studied the ow characteristics of
the diesel spray in a pressurized vessel using PDPA. The axial
velocity of the droplets was studied in the main parts of the spray:
the leading edge, the central part and the trailing edge. From their
analysis, it was found that the temporal distributions of the mean
axial velocity and its rms were constant in the central spray region,
but at the leading edge, they showed a maximum. The radial
distribution of the normalized axial mean velocity was similar to
that of the free gas jet; however, a higher normalized axial mean
velocity value was displayed at the leading edge near the centerline
axis. Moreover, a higher turbulence intensity of the axial velocity
was observed in the leading edge region.
The effect of injection rate shape on the spray characteristics
was also investigated by Benajes et al. [80]. Spray evolution was
produced in three different injection rate shapes: square, ramp and
boot, as shown in Fig. 17. It was found that the square injection rate
always leads to higher spray tip penetration due to the higher
pressure at the SOI, followed by the ramp and boot injection rates.
A study of the cone angle also revealed that there is no inuence of
injection rate shape or injection pressure on the spray cone angle.
By varying the injection pressure and duration for the same injec-
ted quantity, a small difference appears in the spray penetration, in
which higher pressures lead to faster penetration. The combination
of a higher injection pressure with a shorter duration produced
a higher aperture angle at the beginning of the injection process.
Suh et al. [81] studied the spray and combustion characteristics
of a biodiesel/diesel blended fuel in a DI common-rail diesel engine
using visualization and PDPA techniques. The biodiesel fuel that
was used in this work was derived from soybean oil with the
blending ratios of 5% and 20% with diesel fuel. Experiments at
various injection parameters were done in order to study the
inuence of the fuel blending ratio and pilot injection of biodiesel
fuel on the spray characteristics, fuel injection rate, spray tip
penetration and droplet size distribution. The spray characteristics
were studied using CV chamber and the spray visualization was
captured using an ICCDcamera. The atomization characteristic such
as droplet velocity distribution was measured using PDPA tech-
nique. While for combustion characteristics measuring system,
a common-rail diesel engine with four cylinders was installed on an
eddy current dynamometer. It was found that a higher injection
pressure was needed for the biodiesel blended fuel to achieve the
same injection rate at a higher blending ratio. Their study also
revealed that the spray tip penetration of biodiesel was similar to
that of diesel, and the atomization characterization of biodiesel
had a higher Sauter mean diameter and a lower spray velocity.
Combustion analysis also revealed that the peak combustion
pressure of the diesel and biodiesel blended fuel increased with
increasing injection timing. Moreover, the combustion pressure of
the biodiesel blended fuel was higher, as compared to diesel fuel.
Lin and Lin [82] had studied the spray characteristics of methyl
esters fromwaste cooking oil at elevated temperatures. The methyl
ester was treated with two-stage transesterication to lower its
viscosity and to utilize emulsion to reduce the NO
x
pollution. In this
work, methanol was used to enhance the stability and viscosity of
the emulsied fuel. The emulsication also had improved the
methyl ester spray characteristics and decreased the NO
x
emis-
sions, at a temperature above 300

C. The suggested optimal ratio
was waste cooking oil methyl ester 74.5%, methanol 5%, water 20%
and composite surfactant Span-Tween 0.5%.
2.4.1.2. Atomization characteristics. The effect of the injection
driver on the macroscopic and microscopic atomization charac-
teristics of a diesel fuel spray was also investigated by Suh et al. [19]
using a piezo-driven injection system as shown in Fig. 18. The
piezo-driven injection system has a few advantages such as
a shorter injection delay and a fast opening time of the nozzle. It
was found that a shorter injection delay and a faster opening time
were produced by the piezo-driven injection system in comparison
to the solenoid-driven injection system. There was only a small
difference in the tip penetration, but the cone angle of the piezo-
driven system was larger than that of the solenoid-driven injector.
Moreover, the analysis also revealed that the atomization of the
piezo-driven system was better, due to a faster response time and
a higher injection rate.
Fig. 16. Schematic of three-component LDA system used by Chan and Turner [78].
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 735
DME spray characteristics were studied by Suh and Lee [22]
using a common-rail injection system in a diesel engine. This
study focused on the atomization characteristics of DME and
compared experimental and predicted results under various fuel
injection conditions. There are two type of nozzle system that are
used in this work; they are, a single-hole nozzle with 0.3 mm
internal hole diameter (D) and 0.8 mm hole length (L) for spray
visualization and droplet measuring experiments and for fuel
injection rate measurement, different type of nozzle was installed
to improve pressure detection at a lowenergizing duration. In order
to study the injection rate characteristics of diesel and DME fuel,
the injection rate measuring system was applied at various injec-
tion conditions. This measuring system is based on the pressure
variation within a tube lled with the fuel. The pressure in the tube
was kept constant at 3 MPa during fuel injection for the prevention
of DME vapor generation in the tube. Pressure sensor and relief
valve were installed to measure and to keep the constant line
pressure higher than the vapor pressure of DME. In the injection
rate investigation, one thousand continuous injections were aver-
aged for each test case. It was found that the experimental results of
the spray tip penetration showed good agreement with the calcu-
lation results based on empirical equations, except during the
earliest stage of the injection. Atomization analysis revealed that
the DME showed better spray characteristics, as compared to
conventional diesel fuel. Moreover, a shorter fuel injection delay
and a lower maximum rate were produced by DME fuel.
The spray behavior and atomization characteristics of biodiesel
and DME once again were studied by Kim et al. [83] under
increased ambient pressure. They had found that the spray tip
penetration and spray area decreased as the ambient pressure
increased. The atomization performance of the biodiesel was found
to be inferior compared to DME atomization performance.
SME (Soybean oil methyl ester) fuel injection and atomization
characteristics were studied by Park et al. [84] using visualization
and a droplet analyzer system. In this work, they used a single-hole
nozzle 0.3 mm in diameter and 0.8 mm in orice length that was
controlled by a solenoid injector. The injection rates of the diesel
and biodiesel fuels were measured and analyzed using an injection
rate measuring system based on the Bosch method. From their
analysis, it was found that, when the injection pressure increases,
the peak injection rate will increase and advance due to the
increase in the initial injection momentum. The injection rate of
SME is also higher than that of diesel, due to its higher density.
Moreover, the high ambient pressure induced in the combustion
chamber also results in a shorter tip penetration of SME.
Fig. 17. Square, Ramp and Boot type ideal injection rates studied by Benajes et al. [80].
Fig. 18. Comparison of injector driver by Suh et al. [19].
Table 2
Comparison between PIV, LIF, PDPA and Visualization techniques.
Technique Application and parameters
obtained
Disadvantages and
limitations
Visualization Spray and combustion
visualization. Macroscopic
parameter such as spray
geometry, penetration depth,
wall impingement,
and subsequent combustion.
Macroscopic spray
parameters only.
PIV Fuel sprays ow. Instantaneous
whole eld velocities.
Difcult to implement
on dense sprays
LIF Fuel concentration. Liquid
and vapor fuel.
Quenching at high
pressures and
difcult to calibrate.
PDPA Fuel sprays droplet size
measurement. Fuel
droplet size and velocity,
density, volume
ux and time resolved.
Not suitable for high
density sprays.
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 736
Table 3
Summaries of studies of spray and combustion characteristics using optical techniques.
Type of fuel Application Analysis/Comment Reference
Diesel Spray formation, combustion and
soot formation
- Fuel vapor extending beyond liquid core
- Ignition close to the spray tip
- Soot formed in the vapor core region
Bougie et al. [39]
Diesel Evaporating characteristic and
ame distribution
- Extinctions of two wavelengths of ultraviolet
and visible light technique
Yoshizaki et al. [40]
Diesel Macroscopic spray characteristics - Present the way to determine the local spray
angle, cone angle and penetration
Klein-Douwel et al. [41]
Diesel Flames structure of wall-impinging
diesel spray
- Higher OH chemiluminescence were observed
for group-hole nozzles
Gao et al. [42]
Diesel Initial stage of spray development - Tip penetration shows linear dependencies at initial stages
- Tip velocity proportional to t
1/2
Kostas et al. [43]
Gasoline Mixture formation process from two
hole nozzle DISI engine
- Droplet coalescence and penetration increased
but decreased in vapor mass when HAA decreased
- Vaporization and tip penetration rate increased
with increased in injection pressure
Nishida et al. [44]
Natural gas and
hydrogen
Spontaneous ignition and combustion - Natural gas ignition need higher temperature
- Shorter ignition delay for hydrogen
Ishiyama et al. [45]
DME Fuel spray, ame structure, ignition - Shorter ignition delay for DME
- At 1.1 MPa, DME could be ignited at >770 K
Wakai et al. [46]
DME eMethyl
ester blends
Injection and spray characteristics - Shorter injection delay and injection duration,
but higher peak injection duration for
lower blending ratio
- DME evaporated more faster
Bang and Lee [47]
Ethanol-Gasoline
blends
Spray properties - Similar spray pattern for both blended fuel and gasoline
- Better evaporation when the fraction of
ethanol increased at low pressure
- No different in spray penetration at elevated pressure
Gao et al. [48]
Biodiesel blends Spray and combustion development,
effect of injection strategies
- HHR of biodiesel dominated by premixed combustion
- Stronger fuel impingement, longer tip penetration
with increasing biodiesel
Fang et al. [49]
Diesel/Biodiesel Combustion process and NO
x
emissions - NO
x
emissions for biodiesel increased under conventional
combustion, but can be reduced under a certain strategy
Fang et al. [50]
Biodiesel blends (RME) Macroscopic injection parameters - Longer penetration, narrower cone angle, smaller spray
volume due to high viscosity of RME
Desantes et al. [51]
Biodiesel blends (RME) liquid-phase fuel penetration, soot luminosity
and time-resolved intensied imaging
- Fuel penetration increases linearly with blending rates
- Decrease of ignition delay for an increase of RME blend rate
Pastor et al. [52]
Gasoline/multi-
component fuel
Fuel deposition burning in PFI
spark-ignition engine
- Diffusion-controlled near valves due to fuel lm deposition
- soot amount near the valves decreased for advancing ignition
- higher level of soot were observed when the intake valve open
Merola SS and
Vaglieco BM [53]
Gasoline Fuel injection strategy - specic fuel consumption and emissions improved
for the proposed injection strategies
- higher combustion efciency for double injection
Merola SS et al. [54]
Diesel Effect of cavitation on the internal ow - Cavitation leads to an increment in cone angle
- Whole outlet velocity increases when cavitation is initiated
Payri et al. [55]
Water Structure and evolution of cavitation and its
effects on jet pattern
- Proposed conceptual model to explain the development of
asymmetric spray atomization
Ganippa et al. [56]
Diesel/biodiesel Criteria for cavitation inception and its
ow characteristics
- There are 4 regimes in nozzle cavitation ow, that are;
turbulent ow, beginning of cavitation, cavitation
growth and hydraulic ip
- Cavitation decreased with an increase in the Reynolds
number and Weber number
Suh et al. [57]
Diesel Impact of injection pressure - Impact of high injection pressure of biodiesel Delacourt
et al. [58]
Diesel/Biodiesel Spray characteristics under
ultra-high injections
- Longer injection delay and spray tip penetration, but smaller
spray angle, area and volume for biodiesel
- Larger SMD for biodiesel
Wang et al. [59]
Diesel Effect of orice diameter - Lift-off length decreases with decreasing orice diameter
- Total soot incandescence decreases as the orice
diameter decreases
- Normalized ame length decreases as the orice
diameter increases
Pickett and
siebers [60]
Diesel Effect of nozzle geometry - Velocity and mass ow rate strongly inuenced by cavitation Payri et al. [61]
Diesel Effect of nozzle geometry to combustion
characteristics
- CH-radicals was inuenced by spray behavior
- OH-radicals can be considered as measurement of ignition
- Observed the inuence of chamber conditions and
injection pressure
Payri et al. [62]
Diesel Impact of injection angle - Different HRR were observed
- Wall impingement detected on the 70

tip, and produced

2 types of ame
- 150

tip produced weak ame

Fang et al. [63]
Gasoline Swirling injector - Greater air pressure drop at higher injection pressure
- Higher air pressure at nozzle exit at higher fuel temperatures
Moon
et al. [64]
(continued on next page)
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 737
The airow and fuel spray interaction in a variable valve lift
in a PFI engine were studied by Begg et al. [85] using a PDA
technique. The experiment was done in a steady state ow rig
tted with a modied cylinder head, and the intake valve lift
was varied manually. It was found that the mean droplet
diameter was reduced up to 50% for peak valve lifts of less than
3 mm.
2.4.1.3. Droplet velocity characterization. The droplet velocity and
diameter of liqueed butane spray from pintle-type injector were
studied by Koo et al. [86]. Two component phase/Doppler particle
analyzer and traverser were used to obtain the velocity and
diameter of the fuel droplet. The analysis showed that a narrowed
cone angle was observed during the quasi-steady spray period at
0.37 MPa chamber pressure compared to 0.15 MPa. The spray also
Table 3 (continued)
Type of fuel Application Analysis/Comment Reference
Natural gas-
hyrogen blends
Cyclic variation - Cyclic variations initiated at early stage of ame development
- Addition of hydrogen caused a ame kernel
Wang
et al. [65]
n-heptane NO
x
control - Enhanced fuel/air mixing reduced NO
x
- NO
x
reduced with increasing injection pressure
Kitamura
et al. [66]
Propane Turbulent premixed ame - Use of 3-component PIV, very effective for turbulent
combustion process investigation
Stella et al. [67]
Gasoline Spray behavior, Fuel distribution - Fuel spray propagates widely in the downstream direction
- Entropies are higher when the ambient temperature is higher
than fuel evaporation temperature
Lee et al. [67]
DMM, DMC and DME Spray structure, atomization - Larger spray angle, shorter tip penetration for
oxygenated fuel
- Viscosity of oxygenated fuel has signicant
effect on the atomization
Wu et al. [70]
Gasoline GDI ow eld, Stratication effects - Revealed relationship between vorticity and entropy
- Increase in injection timings increases the
homogeneous diffusion rate
Lee and Lee [71]
i-octane, n-dodecane
and n-hexadecane,
and the multi-
component fuels
Spray inner structure, fuel concentration,
fuel evaporation
- Inner structure is very heterogeneous and high
concentration of liquid phase for HI-BP
- Evaporation is promoted by addition LO-BP
Myong et al. [72]
Gasoline Spray structure, liquid and vapor phases - Two regions in the spray, they are; cone and mixing regions
- In the mixing region, vortex ow of the vapor
phase was observed
Choi et al. [74]
Methane Droplet combustion of methane in strained
counter-ow diffusion ames
- Droplet crossing strongly distorted the ame front
- High velocity droplet led to local extinction of the ame
front, and low velocity droplet could ignite within the
ame front and burn on the oxidizer side
Mercier et al. [75]
Gasoline LIF image processing technique for measuring
drop sizes in a pre-swirl spray
- The depth of eld effects in droplet sizing by laser sheet
imaging were effectively measured
Lee and Nishida [76]
Diesel NO concentration in diesel engine - Initial combustion is too fuel-rich for thermal NO formation
- Thermal mechanism is the major pathway for NO formation
Verbiezen et al. [77]
Diesel Measurement of a diesel spray of a normal
size nozzle and a large scale model
- Scaling caused inaccuracy in diesel spray but provided valuable
information about breakup and atomization
Rantanen et al. [35]
e Flow structure - Inclined port creates ow conditions more favorable to
mixing process
- Produces more nearly homogeneous ow structure at TDC
Chan and Turner [78]
Diesel Spray ow characteristics - Temporal distributions were constant in the central pray region
- Higher turbulence intensity of the axial velocity in the
leading edge region
Doudou [79]
Diesel Effect of injection rate shape - Square injection rate leads to higher tip penetration
- No inuence of injection rate to cone angle
- Higher aperture angle obtained by combined a higher injection
pressure with shorter duration
Benajes et al. [80]
Biodiesel and
diesel blended
Spray and combustion, atomization,
SMD, spray velocity
- Higher injection pressure is needed for biodiesel blended fuel
- Similar tip penetration and higher SMD were
observed for biodiesel
Suh et al. [81]
Methyl ester Spray characteristics of methyl ester from
WCO at elevated temperature
- Viscosity of methyl ester was reduced by two-stage
transesterication
- Emulsication improved the spray characteristics and
decreased the NO
x
Lin and Lin [82]
Diesel Effect of injection driver - Shorter injection delay, faster opening time, larger cone
angle and better atomization for piezo-driven injection system
Suh et al. [19]
DME Fuel atomization - DME showed better spray characteristics, shorter injection
delay and lower maximum rate
Suh and Lee [22]
DME Spray behavior and atomization characteristics - Decrease spray tip penetration and cone angle as the ambient
pressure increases
Kim et al. [83]
Soybean oil methyl
ester
Spray structure and atomization - Peak injection rate increase and advance when the injection
pressure increases
- Higher injection rate of SME compared to diesel
Park et al. [84]
Gasoline Effect of variable valve lift - Reduction in mean droplet velocity for peak valve lifts of
less than 3 mm.
Begg et al. [85]
Liqueed butane Droplet velocity and diameter - Narrowed cone angle during quasi-steady spray period
- Spray evaporates faster at 0.37 MPa
Koo et al. [86]
Diesel Spray droplet velocity of convergent nozzles - Higher axial velocity was displayed for nozzles with higher
spray momentum
Payri et al. [87]
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 738
evaporated faster at 0.37 MPa, and the SMDs were similar at spray
edges but nearly doubled in the interior.
Spray droplet velocity characterization for convergent nozzles
was studied by Payri et al. [87] in order to obtain the local speed of
fuel droplets at various points of diesel DI sprays using PDPA tech-
niques. Various orice diameters of convergent nozzles were
investigated, and the experimental measurements were compared
to theoretical results. Their analysis showed that the experimental
results exhibited a high level of agreement with a Gaussian prole.
The experimental data also showedthat the velocity inthe sprayaxis
depends inversely on the axial position. Moreover, a higher axial
velocity was displayed for nozzles with higher spray momentum.
2.5. Summary and recommendations
There are various methods that can be used to capture the spray
and combustion images within the combustion chamber. Different
methods will provide different information. Using visualization
methods, macroscopic parameters canbe determined, andusing the
PIV, LIF and PDPA techniques, information can be found regarding
the microscopic parameters. However, for combustion characteris-
tics studies, visualization technique is very useful in gathering the
information from the combustion chamber. Additionally, three-
component PIV and PDPA techniques can be used to provide better
information on the spray phenomena in IC engines. The comparison
for each technique is shown in Table 2 and the summary of studies
on spray using optical method is shown in Table 3.
The type of alternative fuels that were studied can be classied
into three categories, they are; oxygenated fuel, gaseous fuel and
biodiesel/diesel blended fuel. Most of the oxygenated fuels are
tested with a diesel engine operating system. The oxygenated fuels
were studied fromvarious aspects; the spray structure and motion,
the fuel concentration, the fuel evaporation and its atomization and
were compared to diesel fuel spray and combustion. From their
studies, the comparison between some of oxygenated fuels and
diesel fuel shows; shorter ignition delay and lower temperature for
ignition, shorter spray tip penetration, better spray characteristics
and higher injection rate are needed for higher viscosity oxygen-
ated fuels compared to diesel fuel. Only few studies had focused on
the NO
x
emissions of the oxygenated fuel. While for gaseous fuel,
they were focusing on the visualization of the ignition and
combustion of hydrogen and natural gas. These two types of
gaseous fuel are frequently studied due to their high caloric values
that make themeasily to combust and ease to use in IC engines. The
spray and combustion of LPG (liqueed petroleum gas) and CNG
(compressed natural gas) also need to be studied as they have been
widely used in IC engines. In order to study the ow pattern and
ow velocity of a gaseous fuel, seeding particles are essential to be
added inside the combustion chamber due to ne molecules of the
gaseous fuel, thus making PIV and LIF the most suitable methods
for this kind of measurement. The biodiesel/diesel blended fuel
studies were concentrated on both macroscopic and microscopic
parameters such as their spray and combustion development,
atomization, SMD and spray velocity. Although there were many
reports on the combustion characteristics, but the fundamental
data of the spray and combustion phenomena is still limited
especially on the fuel concentration and evaporation. Further
research can also be conducted on the effect of changing of nozzle
sizes and the effect of injection pressures.
For SI engine combustion, further research can be done on the
effect of the compression ratio, the combustion chamber surface/
volume ratio & geometry, the overall equivalence ratio, the intake
manifold temperature, the intake manifold pressure/throttle posi-
tion, the wall temperature (or rate of engine cooling), the distance
of spark plug fromcylinder axis, the in-cylinder uid motion (swirl,
tumble, squish), the exhaust gas recirculation (EGR) and the spark
timing to the spray and combustion characterization. Previously,
only few studies have been done on SI engine, for example the use
of GDI fuel spray and effect of two-hole nozzle DISI engine, that
leaves a wide scope for researchers in this area.
For CI engine, they were studies on DI engine, but none on IDI
(indirect injection). The injection system in CI engine can be clas-
sied into three categories, they are; mechanical injection systems
(in line, rotary/distributor type), unit injection system (mechanical,
electronic or hydraulically controlled) and CRI (common-rail
injection) systems. Previously, only the effect of piezo-driven
injection was studied by Suh et al. [19].
The stricter emission regulation has also led the researcher to
study the mixing and combustion process of alternative fuels.
Although some of the measurements are done in the CV vessel, it is
also essential to study their emission level that was not studied in
previously. As the detail of spray and combustion data can be
obtained, it will provide the way of controlling the combustion
process, and minimizing the emission level. Any fuel injection
system directly inuences diesel combustion by changes in fuel
injection pressure, injection timing, injected fuel quantity, rate of
fuel injection, quality of spray atomization, possibility of multiple-
injection events and after-injection, dribble & secondary injection.
Previously, only the effect of changing injection parameters such as
injection pressure, injection rate shape and effect of nozzles/orice
number, geometry and angle were studied, hence, leaving a broad
area to be explored in this area of studies. Moreover, with an optical
engine, it is possible to study and control the abnormal combustion
phenomena such as engine knock.
3. Conclusion
A thorough review of published works on spray and combustion
characterization of IC engines using optical measuring technique has
been made. Previous works have been conducted in either modied
engines or CV vessels, with both systems being equipped with an
optical access made of quartz or fused silica. More investigations
have to be carried out on the fundamental of spray and combustion
of fuel in IC engines. Hence, there is a wide scope for researchers in
this area to develop efcient and environmentally clean system.
Alternative fuel, if properly engineered, can be an excellent candi-
date to replaced fossil based fuel, to be applied in automobiles,
stirling engines as well as in domestic and industrial burners. There
are still lots of alternative fuels that can be studied such as a low
heating value producer gas. The fast development in optical
measuring techniques is a promising breakthrough in under-
standing the mixing process, as this provides ways in controlling the
combustion process. There are other optical measurement tech-
niques that can be used such as optical patternator which is suitable
for dense spray as suggested by Sankar [2]. The combustion process
can be improved further by enhancing the mixing process, incor-
porating suitable combustion environment, and changing the
parameters that critically inuence the combustion in a practical
perspective. Moreover, the emissions level studies are essential even
though the experimental works were only done in CV vessels.
Acknowledgement
This research has been nancially supported by the USM-RU-
PGRS provided by UNIVERSITI SAINS MALAYSIA.
References
[1] Ramadhas AS, Jayaraj S, Muraleedharan C. Use of vegetable as IC engine fuels
e A review. Renew Energ 2004;29:727e42.
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 739
[2] Sankar SV, Maher KE, Robart DM, Bachalo WD. Rapid characterization of fuel
atomizers using an optical patternator. J Eng Gas Turb Power 1999;121:409e14.
[3] Su TF, Chang CT, Reitz RD. Effects of injection pressure and nozzle geometry
on spray SMD and D.I. emissions. SAE Paper 952360; 1995.
[4] Nehmer DA, Reitz RD. Measurement of the effect of injection rate and split
injections on diesel engine Soot and NOx emissions. SAE Paper 940668; 1994.
[5] Zhang L. A study of pilot injection in a DI diesel engine. SAE Paper 1999-01-
3493; 1999.
[6] Tanaka T, Ando A, Ishizaka K. Study on pilot injection of DI diesel engine using
common-rail injection system. JSAE 2002;23:297e302.
[7] Onishi S, Hong Jo S, Shoda K, Do Jo P, Kato S. Active thermo-atmosphere
combustion (ATAC) e A new combustion process for internal combustion
engines. SAE Paper 790501; 1979.
[8] Shao J, Yan Y, Greeves G, Smith S. Quantitative characterization of diesel
sprays using digital imaging techniques. Meas Sci Technol 2003;14:1110e6.
[9] Pierpont DA, Reitz RD. The effects of injection pressure and nozzle geometry
on D.I. diesel emissions and performance. SAE Paper 950604; 1995.
[10] McCreath CG, Ber JM. A review of drop size measurement in fuel sprays. Appl
Energ 1976;2(1):3e15.
[11] Jones AR. A review of drop size measurement-the application of techniques to
dense fuel sprays. Prog Energ Combust 1977;3(4):225e34.
[12] Chigier NA. The atomization and burning of liquid fuel sprays. Prog Energ
Combust 1976;2(2):97e114.
[13] Seth B, Aggarwal SK, Sirignano WA. Flame propagation through an air-fuel
spray mixture with transient droplet vaporization. Combust Flame 1980;39
(2):149e68.
[14] Banhawy YE, Whitelaw JH. Experimental study of the interaction between
a fuel spray and surrounding combustion air. Combust Flame 1981;42:253e75.
[15] Yule AJ, Seng CA, Felton PG, Ungut A, Chieger NA. A study of vaporizing fuel
sprays by laser techniques. Combust Flame 1982;44(1e3):71e84.
[16] Nakabe K, Mizutani Y, Hirao T, Fujioka H. An experimental study on detailed
ame structure of liquid fuel sprays with and without gaseous fuel. Combust
Flame 1991;1-2:3e14.
[17] Presser C, Gupta AK, Avedisian CT, Semerjian HG. Fuel property effects on the
structure of spray ames. Sym Int Combust Proc 1991;23(1):1361e7.
[18] Senda J, Kobayashi M, Tanabe Y, Fujimoto H. Visualization and quantitative
analysis of the fuel vapor concentration in diesel spray. JSAE Review 1994;15
(2):149e56.
[19] Suh HK, Park SW, Lee CS. Effect of piezo-driven injection system on the
macroscopic and microscopic atomization characteristics of diesel fuel spray.
Fuel 2007;86:2833e45.
[20] Kawano D, Senda J, Kawakimi K, Shimada A, Fujimoto H. Fuel design concept
for low emission in engine system 3
rd
report: analysis of spray characteristics
for mixed fuels. SAE Paper 2002-01-0220; 2002.
[21] Kawanabe H, Kondo C, Kohori S, Shioji M. Simultaneous measurements of
velocity and scalar elds in a turbulent jet using PIV and LIF. JSME 2008;74
(742):1411e8.
[22] Suh HK, Lee CS. Experimental and analytical study on the spray characteristics
of dimethyl ether (DME) and diesel fuels within a common-rail injection
system in a diesel engine. Fuel 2008;87:925e32.
[23] Dent JC. Basis for the comparison of various experimental methods for
studying spray penetration. SAE Paper 710571; 1971.
[24] Hiroyasu H, Arai M. Structures of fuels sprays in Diesel engines. SAE Paper
1990;900475.
[25] Bosch W. The fuel rate indicator: a new measuring instrument for display of
the characteristics of individual injection. SAE Paper; 1966:660e749.
[26] Heywood JB. Internal combustion engine fundamental. New York: McGraw-
Hill International Editions; 1988.
[27] Dan T, Senda J, Fujimoto H. Organized structure and motion in diesel spray.
SAE Paper 970641; 1997.
[28] Kosaka H, Kamimoto T. A study on the structure of transients sprays. JSME
1999;65(640):272e7.
[29] Yuyama R, Chikahisa T, Kikuta K, Hishinuma Y. Entropy analysis of micro-
scopic diffusion phenomena in diesel sprays. COMODIA; 2001:542e50.
[30] Zainal ZA, Soid SN. Modeling of combustion characteristic in diesel engine
fueled by producer gas. SAE Paper 2007-01-2130; 2007.
[31] Ferguson CR, Kirkpatrick AT. Internal combustion engines. 2nd ed. New York:
John Wiley & Sons Inc; 2001.
[32] Palopski T. The use of dimensional analysis and model experiments in the
research of fuel injection (in Finnish). Helsinki University of technology,
Laboratory of energy engineering and environmental protection, Espoo.
[33] Walzel P. Advantages and limits in large scale modeling of atomizers. ICLASS;
1982:187e94.
[34] Soteriou C, Andrews R, Smith M. Direct injection diesel sprays and the effect of
cavitation and hydraulic ip on atomization. SAE Paper 950080; 1995.
[35] RantanenP, ValkonenA, Cronhjort. Measurement of a diesel spray witha normal
size nozzle and a large-scale model. Int J Heat Fluid Fl 1999;20:545e51.
[36] Roisman IV, Araneo L, Tropea C. Effect of ambient pressure on penetration of
a diesel spray. Int J Multiphas Flow 2007;33:904e20.
[37] Chigier N. Optical imaging of sprays. Prog Energ Combust 1991;17(3):211e62.
[38] Zhao H, Ladommatos N. Optical Diagnostics for in-cylinder mixture formation
measurement in IC engines. Prog Energ Combust 1998;24:297e336.
[39] Bougie B, Tulej M, Dreier T, Dam NJ, Ter Meulen JJ, Gerber T. Optical diag-
nostics of diesel spray injections and combustion in a high-pressure high-
temperature cell. Appl Phys B-Lasers O 2005;80:1039e45.
[40] Yoshizaki T, Yamamoto O, Nishida K, Hiroyasu H, Suzuki M. Evaporating spray
and ame distributions in the combustion chamber of a direct injection diesel
engine. JSAE Rev 1997;18(3):219e24.
[41] Klein-Douwel RJH, Frijters PJM, Somers LMT, de Boer WA, Baert RSG.
Macroscopic diesel fuel spray shadowgraphy using high speed digital imaging
in a high pressure cell. Fuel 2007;86:1994e2007.
[42] Gao J, Moon S, Zhang Y, Nishida K, Matsumoto Y. Flame structure of wall-
impinging diesel fuel sprays injected by group-hole nozzle. Combust Flame
2009;156(6):1263e77.
[43] Kostas J, Honnery D, Soria J. Time resolved measurements of the initial stages
of fuel spray penetration. Fuel 2009;88(11):2225e37.
[44] Nishida K, Tian J, Sumoto Y, Long W, Sato K, Yamakawa M. An experimental
and numerical study on sprays injected from two-hole nozzles for DISI
engines. Fuel 2009;88(9):1634e42.
[45] Ishiyama T, Shioji M, Ihara T, Inoue T, Takada N. Characteristics of sponta-
neous ignition and combustion in unsteady high speed gaseous fuel jets. SAE
Paper 2003-01-1922; 2003.
[46] Wakai K, Nishida K, Yoshizaki T, Hiroyasu H. Spray and ignition character-
istics of di-methyl ether injected by a DI diesel injector. COMODIA;
1998:537e42.
[47] Bang SH, Lee CS. Fuel injection characteristics and spray behavior of DME
blended with methyl ester derived from soybean oil. Fuel 2010;89:797e800.
[48] Gao J, Jiang D, Huang Z. Spray properties of alternative fuels: a comparative
analysis of ethanol-gasoline blends and gasoline. Fuel 2007;86:1646e50.
[49] Fang T, Lin YC, Foong TM, Lee CF. Biodiesel combustion in an optical HSDI
diesel engine under low load premixed combustion conditions. Fuel 2009;88
(11):2154e62.
[50] FangT, LeeCF. Bio-diesel effects oncombustionprocesses inanHSDI diesel engine
using advanced injection strategies. P Combust Inst 2009;32(2):2785e92.
[51] Desantes JM, Payri R, Garcia A, Manin J. Experimental study of biodiesel
blends effects on diesel injection processes. Energ Fuel 2009;23:3227e35.
[52] Pastor JV, Payri R, Gimeno J, Nerva JG. Experimental study on RME blends:
liquid-phase fuel penetration, chemiluminescence, and soot luminosity in
diesel-like conditions. Energ Fuel 2009;23:5899e915.
[53] Merola SS, VagliecoBM. Optical investigations of fuel depositionburninginported
fuel injection (PFI) spark-ignition (SI) engine. Energy 2009;34(12):2108e15.
[54] Merola SS, Sementa P, Tornatore C, Vaglieco BM. Effect of the fuel injection
strategy on the combustion process in a PFI boosted spark-ignition engine.
Energy 2010;35(2):1094e100.
[55] Payri F, Bermudez V, Payri P, Salvador FJ. The inuence of cavitation on the
internal ow and the spray characteristics in diesel injection nozzles. Fuel
2004;83:419e31.
[56] Ganippa LC, Bark G, Andersson S, Chomiak J. Cavitation: a contributory factor
in the transition from symmetric to asymmetric jets in cross-ow nozzles. Exp
Fluids 2004;36:627e34.
[57] Suh HK, Park SH, Lee CS. Experimental investigation of nozzle cavitating ow
characteristics for diesel andbiodiesel fuels. Int J Automot Techn2008;9:217e24.
[58] Delacourt E, Desmet B, Besson B. Characterisation of very high pressure diesel
sprays using digital imaging techniques. Fuel 2005;85:859e67.
[59] Wang X, Huang Z, Kuti OA, Zhang W, Nishida K. Experimental and analytical
study of biodiesel and diesel spray characteristics under ultra-high injection
pressure. Int J Heat Fluid Fl 2010;31(4):656e66.
[60] Pickett LM, Siebers DL. Orice diameter effects on diesel fuel jet ame
structure. J Eng Gas Turb Power 2005;127:187e96.
[61] Payri R, Salvador FJ, GimenoJ, ZapataLD. Diesel nozzlegeometryinuenceonspray
liquid-phase fuel penetration in evaporative conditions. Fuel 2008;87:1165e76.
[62] Payri R, Salvador FJ, Gimeno J, de la Morena J. Effects of nozzle geometry ondirect
injection diesel engine combustion process. Appl ThermEng 2009;29:2051e60.
[63] Fang T, Coverdill RE, Lee CF, White RA. Effects of injection angles on
combustion processes using multiple injection strategies in an HSDI diesel
engine. Fuel 2008;87:3232e9.
[64] Moon S, Abo-Serie E, Bae C. Air ow and pressure inside a pressure-swirl
spray and their effects on spray development. Exp Thermal Fluid Sci
2009;33:222e31.
[65] Wang J, Huang Z, Miao H, Wang X, Jiang D. Study of cyclic variations of direct-
injection combustion fueled with natural gas-hydrogen blends using
a constant volume vessel. Int J Hydrogen Energ 2008;33:7580e91.
[66] Kitamura Y, Kee SS, Mohammadi A, Ishiyama T, Shioji M. Study on NOx
control in direct-injection PCCI combustion-fundamental investigation using
a constant-volume vessel. SAE Paper; 2006. 2006-01-0919.
[67] Lee KH, Lee CH, Lee CS. An experimental study on the spray behavior and fuel
distribution of GDI injectors using the entropy analysis and PIV method. Fuel
2004;83:971e80.
[68] Kompenhans J, Khler C. Particle image velocimetry e An advanced experi-
mental tool for the investigation of turbulent ow elds. Els Ser Therm Fluid
2002;5:79e94.
[69] Stella A, Guj G, Kompenhans J, Richard H, Raffel M. Three-component particle
image velocimetry measurements in premixed ames. Aerosp Sci Technol
2001;5:357e64.
[70] Wu Z, Zhu Z, Huang Z. An experimental study on the spray structure of
oxygenated fuel using laser-based visualization and particle image velocim-
etry. Fuel 2006;85:1458e64.
[71] Lee KH, Lee CH. Characterization of the ow eld and stratication effects of
fuel spray in a visualization engine using DPIV and entropy analysis. Exp
Therm Fluid Sci 2007;31:579e92.
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 740
[72] Myong KJ, Suzuki H, Senda J, Fujimoto H. Spray inner structure of evaporating
multi-component fuel. Fuel 2008;87:202e10.
[73] Desantes JM, Pastor JV, Pastor JM, Julia JE. Limitations on the use of the planar
laser induced exciplex uorescence technique in diesel sprays. Fuel
2005;84:2301e15.
[74] Choi DS, Choi GM, Kim DJ. Spray structures and vaporizing characteristics of
a GDI fuel spray. KSME Int J 2002;16:999e1008.
[75] Mercier X, Orain M, Grisch F. Investigation of droplet combustion in strained
counterow diffusion ames using planar laser-induced uorescence. Appl
Phys B-Lasers O 2007;88:151e60.
[76] Lee JK, Nishida K. Development of an LIF image processing technique for
measuring drop sizes in a pre-swirl spray. Int J Automot Techn
2008;9:381e90.
[77] Verbiezen K, Donkerbroek AJ, Klein-Douwel RJH, Vliet APV, Frijters PJM,
Seykens XLJ, et al. Diesel combustion: in-cylinder NO concentrations in rela-
tion to injection timing. Combust Flame 2007;151:333e46.
[78] Chan VSS, Turner JT. Velocity measurement inside a motored internal
combustion engine using three-component laser Doppler anemometry. Opt
Laser Technol 2000;32:557e66.
[79] Doudou A. Turbulent ow study of an isothermal diesel injected by a common
rail system. Fuel 2005;84:287e98.
[80] Benajes J, Payri R, Molina S, Soare V. Investigation of the inuence of injection
rate shaping on the spray characteristics in a diesel common-rail system
equipped with a piston amplier. J Fluids Eng-T ASME 2005;127:1102e10.
[81] Suh HK, Roh HG, Lee CS. Spray and combustion characteristics of biodiesel/
diesel blended fuel in a direct injection common-rail diesel engine. J Eng Gas
Turb Power 2008;130:032807e9.
[82] Lin YS, Lin HP. Study on the spray characteristics of methyl esters from waste
cooking oil at elevated temperature. Renew Energ 2010;35:1900e7.
[83] Kim JK, Park SH, Lee CS. A study on the macroscopic spray behavior and
atomization characteristics of biodiesel and dimethyl ether sprays under
increased ambient pressure. Fuel Process Technol 2010;91(3):354e63.
[84] Park SH, Kim HJ, Suh HK, Lee CS. A study on the fuel injection and atomization
characteristics of soybean oil methyl ester (SME). Int J Heat Fluid
2009;30:108e16.
[85] Begg SM, Hindle MP, Cowell T, Heikal MR. Low intake valve lift in a port fuel-
injected engine. Energy 2009;34(12):2042e50.
[86] Koo JY, Chung HC, Shakal JS, Goto S. Characteristics of a liqueed butane spray.
KSME Int J 1998;12:112e22.
[87] Payri R, Tormos B, Salvador FJ, Araneo L. Spray droplet and velocity charac-
terization for convergent nozzles with three different diameters. Fuel
2008;87:3176e82.
S.N. Soid, Z.A. Zainal / Energy 36 (2011) 724e741 741